We build our investment strategies for long-term investors in four steps.
First, we set target levels for risk, from cautious to adventurous investors. Second, we define the asset classes we use. At a broad level, these are shares, bonds, cash and alternatives. Then we forecast risk and return for each asset class, including how they interact with each other. And lastly, we combine asset classes to build portfolios that offer the highest expected return for each level of risk.
Here, we focus on how we forecast returns. There are four ingredients:
The returns implicit in market prices
Investment theory
Consensus views
Historical returns
This article isn’t personal advice. Remember, investments and any income from them can rise and fall in value, so you could get back less than you invest. Past performance isn’t a guide to the future. If you’re not sure if an investment’s right for you, ask for financial advice.
Market-implied returns
There are over 70 UK government bonds in issue today. At the time of writing, the shortest duration bond is due to be repaid in 11 days while the longest expires in 2073.
Comparing, for example, the yield on a one-year bond with a two-year one allows us to calculate what the market expects the yield on a one-year bond to be in a year’s time. Comparing all 70 gives us a comprehensive picture of market expectations.
We can carry out a similar exercise for government bonds in other countries, which also allows us to extract implied forecasts for currencies. Adding analysis of inflation-linked bonds, corporate bonds, options and derivatives markets gives us forecasts for a long list of markets – cash, government fixed-income bonds, government inflation-linked bonds, corporate bonds, exchange rates, volatility and inflation.
Investment theory
There’s no equivalent easy way of extracting expected returns for shares from market pricing. Instead, we turn to investment theory.
This gives us two possible approaches.
The first takes a top-down view, looking at global equities as a whole. We should expect a higher return on shares than bonds, to compensate us for the additional risks. We assume a premium of 5% per annum, informed by a range of academic studies. We adjust forecasts for individual markets based on their relative riskiness.
Therefore, for example, we expect a higher return for more volatile smaller companies and a lower return for more stable infrastructure shares.
The second approach builds forecasts from the bottom up. For example, the high valuation of the US equity market translates into a lower expected return, all else being equal. Investors must then decide if the higher valuation will be justified, either because of its more dynamic economy or because it’s home to the companies expected to benefit most from the boom in artificial intelligence.
Consensus views
Government bonds come in two varieties – fixed interest and inflation linked. By comparing the two, we can calculate the inflation rate expected by the market. However, inflation-linked bonds are a particularly good insurance policy for pension providers and insurance companies that want to hedge their long-term liabilities. Like all insurance policies, there’s a price to pay – and this means the implied inflation rate isn’t a pure measure of investor expectations.
Surveys of professional economists provide an alternative measure of long-term inflation expectations, avoiding this distortion. The Federal Reserve Bank of Philadelphia offers a publicly available measure of US inflation expectations. For other countries, forecasts are available from commercial data providers.
Historical returns
Past performance is not a guide to the future. However, a study of long run returns can still provide useful context for long-term investors.
The UBS Global Investment Returns Yearbook documents the returns on stocks, bonds, cash and currencies since 1900 for 23 countries, alongside shorter histories for over 60 more. This study allows us to understand how our current forecasts compare to realised 10-year returns in many different economic and political environments.
For some asset classes, like commodity futures, we find no robust theoretical approach to predicting returns, nor a market implied forecast, nor a consensus view. Here, historical returns provide us at least some sense of scale for likely risk and return going forward.
We also lean on history when forecasting how markets interact with each other. We consider the longest history possible but put more emphasis on more recent data.
Be humble
If you invest in shares and bonds, you are implicitly making a forecast – that these asset classes are expected to outperform cash over the long term. We agree! But we know that some decades are better than others.
Sometimes diversification is effective, other times less so. We consider a wide range of outcomes when building our portfolios. We stress test them against historical downturns and forward-looking scenarios to better understand risks.
A forecast of the expected return of your portfolio can help you plan for a better financial future. But do make sure that your plans cover the risks embedded in those forecasts.